EDITOR’S PROOF
272 H.D. Clarke et al.
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
push the UK into recession. A middle category allowed respondents to say they did
not know which option to choose. A fifth question asked respondents about whether
they favored more or fewer services from the government, with a “don’t know” op-
tion as well. Responses to the five questions were rescaled to produce high scores
when the respondent thought that cuts were needed to solve the UK’s economic
problems, whereas low scores indicated that the respondent believed that the cuts
would be harmful. A principal components exploratory factor analysis indicated
that a single factor structures answers to the five questions, and we use factor scores
produced by this analysis as the dependent variable. Given the continuous nature of
the dependent variable, our model of attitudes towards the cuts was estimated with
ordinary least squares regression.
Predictor variables included demographic measures for gender, age, ethnicity,
education and income bands. Gender was a 0–1 dummy variable and age was mea-
sured in years. We expected that men, who traditionally have less responsibility
for child and family care, would be more likely to favor the cuts. For age, we en-
tertained two possibilities; older people might be more conservative and favor the
cuts, or they might recognize the vulnerability of old age and oppose them. We also
computed a new variable, the square of a respondent’s age, in an effort to capture
possible curvilinear effects of age. Ethnicity was dichotomized into “white British”
and other ethnicity and race identifications, with minorities scored as 1 and “white
British” as 0. As a vulnerable social group, we expected non-whites to be opposed
to the cuts. Income was measured in 14 bands. As income increased, we anticipated
that support for the cuts to increase, but education proved to be a trickier prediction.
Education often correlates with income, but the more highly educated also might
be more sympathetic to the need for an extensive set of publicly funded social pro-
grams.
The model also included dummy variables for Scotland and Wales to determine
if regional differences emerged. Scotland in particular is considered to be consid-
erably more left in its ideological proclivities than is the UK as a whole, and we
hypothesized that being a resident of Scotland would produce a negative coefficient
in the multivariate analysis. We made no such prediction for Wales.
We also included a dummy variable to evaluate the effects of workforce status
and vulnerability, combining short- and long-term unemployed into a single cate-
gory with the permanently disabled and ill and those with long-term caregiver re-
sponsibilities. We predicted that those who were unemployed would find the pub-
lic spending cuts harsh, both because benefits were reduced and because spending
cuts meant fewer opportunities for job seekers. Similarly we predicted that the sick,
disabled and caregivers would manifest less support for the cuts than would other
people.
As elsewhere, the economy is a major concern for most citizens of the UK. Our
model of attitudes towards the cuts contains a predictor variable measuring cog-
nitive evaluations of national and personal economic evaluations, constructed via
an exploratory factor analysis. The BES routinely measures economic evaluations
with four questions on five-point Likert scales. The questions elicit sociotropic and
egocentric evaluations both retrospectively and prospectively. The factor analysis of
